Content area
Full Text
1. Introduction
With the rapid advancements in artificial intelligence technology, mobile robots have increasingly taken on the role of human operators in various practical operations, offering improved efficiency and safety. Consequently, these robotic systems, encompassing sensors, remote controls, automatic controllers and other mobile capabilities, have become integral components in an array of application scenarios. State estimation and localization in unknown environments have emerged as prominent research areas in the domain of mobile robotics, with SLAM serving as a focal point. Compared to cameras, the utilization of LiDAR technology provides notable advantages, as it is unaffected by ambient light and texture, allowing for highly accurate and efficient distance measurements. The LiDAR-based SLAM system has been extensively developed in the fields of automated driving (Zhang et al., 2022; Badue et al., 2021), mobile robots, forestry surveying (Tao et al., 2021), urban surveying and mapping (Liu et al., 2017).
Tee provided a comprehensive analysis and comparison of several popular open-source implementations of 2D LiDAR-based SLAM (Tee and Han, 2021). However, the investigation solely focused on 2D LiDAR-based SLAM techniques, with no mention of their 3D counterparts. Bresson examined the application of LiDAR-based SLAM specifically within the context of the grand challenge of autonomous driving (Bresson et al., 2017). Notably, Xu et al. (2022b) presented an in-depth exploration of the development of multi-sensor fusion positioning, with meticulous attention given to the evaluation of both loosely coupled and tightly coupled systems. This paper presents a novel approach to reviewing the literature on LiDAR-based SLAM by focusing on the application of different types and configurations of LiDAR. This paper offers a significant contribution as a reference for researchers and engineers seeking to gain insight into the wide-ranging applications of different LiDAR types and configurations, distinguishing itself from previous review studies.
The remainder of this paper is organized as follows. Section 2 provides an anatomy of a LiDAR-based SLAM system. In Section 3, the related work of LiDAR-based SLAM systems is reviewed in three segments based on LiDAR types and configurations. Section 4 proposes several new frontiers in LiDAR-based SLAM. Finally, Section 5 concludes this paper.
2. Anatomy of a light detection and ranging-based simultaneous localization and mapping system
2.1 Historical perspective
Smith and Cheeseman (1986) united...